Discover GPU acceleration for JavaScript in this JSConf.Asia 2016 conference talk. Explore the creation of GPU.JS, a project born from a 24-hour hackathon that enables running JavaScript code on GPUs. Learn about the technology behind GPU.JS, including its API, compilation to GLSL, and the SPMD computation model. Follow along with practical examples such as matrix multiplication and real-time ray tracing. Gain insights into the project's development, its potential applications, and limitations. Perfect for JavaScript developers interested in leveraging GPU power for performance optimization and parallel computing enthusiasts looking to expand their knowledge.
Overview
Syllabus
Intro
Fazli Sapuan
Eugene Cheah Pico Creator
The Story Behind GPU.JS
And it was a Nintendo fun start
Literally under 24 hours
Lesson learnt?
All in one convenient API package
Getting Started with GPU.JS
Technology Overview
Compiling to GLSL
Turning our inputs into textures
Writing our output
Running the Program
The Magic in 3 Sentences
Caveats (cont'd...)
SPMD Computation Model
Example: Matrix Multiplication
Demo: Ray Tracer
Real-Time Ray Tracing
Summary
Interested to find out more?
Extra: Join Hack&Roll 2017 (students only)
Taught by
JSConf
